So Unstable and driving me mad!!

[Mark Haney]

Carl Friis-Hansen wrote:
> Mark Haney wrote:
>> It's funny, I rarely ever see a system completely bomb because of funky 
>> memory anymore.  It always exhibits bizarre and unstable behaviour 
>> rather than simply not booting or a panicked kernel.  Some of that might 
>> be due to the way the memory bus is designed now.
>>
> Mark, you answered yourself why Karl was not completely wrong. The thing 
> is that modern machines are a lot more stable and that even though they 
> are almost 10,000 times faster today compared to the 4.75MHz IBM. The 
> reason is simple: Computers consist of fever and fever ICs and of 
> shorter and shorter inter connections. The manufactures of the VLSIs and 
> CPUs are doing a very good design job with a very low failure rate. 
> Because the circuit board is approaching elimination, the product of 
> speed and complexity has increased and the failure rate decreased.
> Yes, an old well proven machine will probably continue work very well, 
> whatever you put on it, but there still is some truth in Karl's earlier 
> post.
> 

I disagree.  I do not believe modern machines are more stable. Not one 
bit.  I see more stability issues with newer systems than older ones. 10 
years ago I rarely saw memory problems like I see now.  And believe me I 
have built well over a thousand systems in the last 12 years or so.

I submit that the more complex a system is, the higher probability of 
encountering a problem.  I would like to see numbers to back up your 
statement that complexity has increased and failures decreased.  This 
comes on the heels of the NVidia GPU problems, Seagate's 1.5TB failure 
rates, etc.  Truthfully, we can argue this point ad infinitum, and never 
see a true answer.

I simply state this Karl's assumption that the OP's system was really 
old to be complete fallacy.  I have a system at home that I built 10 
years ago (Dual Pentium 300) that hasn't seen a problem or reboot since 
late 2002.

He assumes 'old' is 'worn out' and that's not always the case.  It was a 
complete waste of bandwidth to post what he did and offer NO suggestion 
on how to debug the problem.


-- 
Frustra laborant quotquot se calculationibus fatigant pro inventione 
quadraturae circuli

Mark Haney
Sr. Systems Administrator
ERC Broadband
(828) 350-2415

Call (866) ERC-7110 for after hours support